Github user yanboliang commented on a diff in the pull request:
https://github.com/apache/spark/pull/11136#discussion_r53592447
--- Diff:
mllib/src/main/scala/org/apache/spark/ml/regression/GeneralizedLinearRegression.scala
---
@@ -0,0 +1,547 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements. See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+package org.apache.spark.ml.regression
+
+import breeze.stats.distributions.{Gaussian => GD}
+
+import org.apache.spark.{Logging, SparkException}
+import org.apache.spark.annotation.{Experimental, Since}
+import org.apache.spark.ml.PredictorParams
+import org.apache.spark.ml.feature.Instance
+import org.apache.spark.ml.optim._
+import org.apache.spark.ml.param._
+import org.apache.spark.ml.param.shared._
+import org.apache.spark.ml.util.Identifiable
+import org.apache.spark.mllib.linalg.{BLAS, Vector}
+import org.apache.spark.rdd.RDD
+import org.apache.spark.sql.{DataFrame, Row}
+import org.apache.spark.sql.functions._
+
+/**
+ * Params for Generalized Linear Regression.
+ */
+private[regression] trait GeneralizedLinearRegressionParams extends
PredictorParams
+ with HasFitIntercept with HasMaxIter with HasTol with HasRegParam with
HasWeightCol
+ with HasSolver with Logging {
+
+ /**
+ * Param for the name of family which is a description of the error
distribution
+ * to be used in the model.
+ * Supported options: "gaussian", "binomial", "poisson" and "gamma".
+ * @group param
+ */
+ @Since("2.0.0")
+ final val family: Param[String] = new Param(this, "family",
+ "the name of family which is a description of the error distribution
to be used in the model",
+
ParamValidators.inArray[String](GeneralizedLinearRegression.supportedFamilies.toArray))
+
+ /** @group getParam */
+ @Since("2.0.0")
+ def getFamily: String = $(family)
+
+ /**
+ * Param for the name of the model link function.
+ * Supported options: "identity", "log", "inverse", "logit", "probit",
"cloglog" and "sqrt".
+ * @group param
+ */
+ @Since("2.0.0")
+ final val link: Param[String] = new Param(this, "link", "the name of the
model link function",
+
ParamValidators.inArray[String](GeneralizedLinearRegression.supportedLinks.toArray))
+
+ /** @group getParam */
+ @Since("2.0.0")
+ def getLink: String = $(link)
+
+ @Since("2.0.0")
+ override def validateParams(): Unit = {
+ if (isDefined(link)) {
+
require(GeneralizedLinearRegression.supportedFamilyLinkPairs.contains($(family)
-> $(link)),
+ s"Generalized Linear Regression with ${$(family)} family does not
support ${$(link)} " +
+ s"link function.")
+ }
+ }
+}
+
+/**
+ * :: Experimental ::
+ *
+ * Fit a Generalized Linear Model
([[https://en.wikipedia.org/wiki/Generalized_linear_model]])
+ * specified by giving a symbolic description of the linear predictor and
+ * a description of the error distribution.
+ */
+@Experimental
+@Since("2.0.0")
+class GeneralizedLinearRegression @Since("2.0.0") (@Since("2.0.0")
override val uid: String)
+ extends Regressor[Vector, GeneralizedLinearRegression,
GeneralizedLinearRegressionModel]
+ with GeneralizedLinearRegressionParams with Logging {
+
+ @Since("2.0.0")
+ def this() = this(Identifiable.randomUID("genLinReg"))
+
+ /**
+ * Set the name of family which is a description of the error
distribution
+ * to be used in the model.
+ * @group setParam
+ */
+ @Since("2.0.0")
+ def setFamily(value: String): this.type = set(family, value)
+
+ /**
+ * Set the name of the model link function.
+ * @group setParam
+ */
+ @Since("2.0.0")
+ def setLink(value: String): this.type = set(link, value)
+
+ /**
+ * Set if we should fit the intercept.
+ * Default is true.
+ * @group setParam
+ */
+ @Since("2.0.0")
+ def setFitIntercept(value: Boolean): this.type = set(fitIntercept, value)
+ setDefault(fitIntercept -> true)
+
+ /**
+ * Set the maximum number of iterations.
+ * Default is 100.
+ * @group setParam
+ */
+ @Since("2.0.0")
+ def setMaxIter(value: Int): this.type = set(maxIter, value)
+ setDefault(maxIter -> 100)
+
+ /**
+ * Set the convergence tolerance of iterations.
+ * Smaller value will lead to higher accuracy with the cost of more
iterations.
+ * Default is 1E-6.
+ * @group setParam
+ */
+ @Since("2.0.0")
+ def setTol(value: Double): this.type = set(tol, value)
+ setDefault(tol -> 1E-6)
+
+ /**
+ * Set the regularization parameter.
+ * Default is 0.0.
+ * @group setParam
+ */
+ @Since("2.0.0")
+ def setRegParam(value: Double): this.type = set(regParam, value)
+ setDefault(regParam -> 0.0)
+
+ /**
+ * Whether to over-/under-sample training instances according to the
given weights in weightCol.
+ * If empty, all instances are treated equally (weight 1.0).
+ * Default is empty, so all instances have weight one.
+ * @group setParam
+ */
+ @Since("2.0.0")
+ def setWeightCol(value: String): this.type = set(weightCol, value)
+ setDefault(weightCol -> "")
+
+ /**
+ * Set the solver algorithm used for optimization.
+ * Currently only support "irls" which is also the default solver.
+ * @group setParam
+ */
+ @Since("2.0.0")
+ def setSolver(value: String): this.type = set(solver, value)
+ setDefault(solver -> "irls")
+
+ override protected def train(dataset: DataFrame):
GeneralizedLinearRegressionModel = {
+ val familyLink = $(family) match {
+ case "gaussian" => if (isDefined(link)) Gaussian($(link)) else
Gaussian("identity")
+ case "binomial" => if (isDefined(link)) Binomial($(link)) else
Binomial("logit")
+ case "poisson" => if (isDefined(link)) Poisson($(link)) else
Poisson("log")
+ case "gamma" => if (isDefined(link)) Gamma($(link)) else
Gamma("inverse")
+ }
+
+ val numFeatures = dataset.select(col($(featuresCol))).limit(1).map {
+ case Row(features: Vector) => features.size
+ }.first()
+ if (numFeatures > 4096) {
+ val msg = "Currently, GeneralizedLinearRegression only supports
number of features" +
+ s" <= 4096. Found $numFeatures in the input dataset."
+ throw new SparkException(msg)
+ }
+
+ val w = if ($(weightCol).isEmpty) lit(1.0) else col($(weightCol))
+ val instances: RDD[Instance] = dataset.select(
+ col($(labelCol)), w, col($(featuresCol))).map {
+ case Row(label: Double, weight: Double, features: Vector) =>
+ Instance(label, weight, features)
+ }
+
+ if ($(family) == "gaussian" && $(link) == "identity") {
+ val optimizer = new WeightedLeastSquares($(fitIntercept),
$(regParam),
+ standardizeFeatures = true, standardizeLabel = true)
+ val wlsModel = optimizer.fit(instances)
+ val model = copyValues(new GeneralizedLinearRegressionModel(uid,
+ wlsModel.coefficients, wlsModel.intercept).setParent(this))
+ return model
+ }
+
+ val newInstances = instances.map { instance =>
+ val mu = familyLink.initialize(instance.label, instance.weight)
+ val eta = familyLink.predict(mu)
+ Instance(eta, instance.weight, instance.features)
+ }
+
+ val initialModel = new WeightedLeastSquares($(fitIntercept),
$(regParam),
+ standardizeFeatures = true, standardizeLabel =
true).fit(newInstances)
+
+ val reweightFunc: (Instance, WeightedLeastSquaresModel) => (Double,
Double) = {
+ (instance: Instance, model: WeightedLeastSquaresModel) => {
+ val eta = model.predict(instance.features)
+ val mu = familyLink.fitted(eta)
+ val z = familyLink.adjusted(instance.label, mu, eta)
+ val w = familyLink.weights(mu) * instance.weight
+ (z, w)
+ }
+ }
+
+ val optimizer = new IterativelyReweightedLeastSquares(initialModel,
reweightFunc,
+ $(fitIntercept), $(regParam), $(maxIter), $(tol))
+
+ val irlsModel = optimizer.fit(instances)
+
+ val model = copyValues(new GeneralizedLinearRegressionModel(uid,
+ irlsModel.coefficients, irlsModel.intercept).setParent(this))
+ model
+ }
+
+ @Since("2.0.0")
+ override def copy(extra: ParamMap): GeneralizedLinearRegression =
defaultCopy(extra)
+}
+
+@Since("2.0.0")
+object GeneralizedLinearRegression {
+
+ /** Set of families that GeneralizedLinearRegression supports */
+ private[ml] val supportedFamilies = Set("gaussian", "binomial",
"poisson", "gamma")
+
+ /** Set of links that GeneralizedLinearRegression supports */
+ private[ml] val supportedLinks = Set("identity", "log", "inverse",
"logit", "probit",
+ "cloglog", "sqrt")
+
+ /** Set of family and link pairs that GeneralizedLinearRegression
supports */
+ private[ml] val supportedFamilyLinkPairs = Set(
+ "gaussian" -> "identity", "gaussian" -> "log", "gaussian" -> "inverse",
+ "binomial" -> "logit", "binomial" -> "probit", "binomial" -> "cloglog",
+ "poisson" -> "log", "poisson" -> "identity", "poisson" -> "sqrt",
+ "gamma" -> "inverse", "gamma" -> "identity", "gamma" -> "log"
+ )
+}
+
+/**
+ * :: Experimental ::
+ * Model produced by [[GeneralizedLinearRegression]].
+ */
+@Experimental
+@Since("2.0.0")
+class GeneralizedLinearRegressionModel private[ml] (
+ @Since("2.0.0") override val uid: String,
+ @Since("2.0.0") val coefficients: Vector,
+ @Since("2.0.0") val intercept: Double)
+ extends RegressionModel[Vector, GeneralizedLinearRegressionModel]
+ with GeneralizedLinearRegressionParams {
+
+ private lazy val familyLink = $(family) match {
+ case "gaussian" => if (isDefined(link)) Gaussian($(link)) else
Gaussian("identity")
+ case "binomial" => if (isDefined(link)) Binomial($(link)) else
Binomial("logit")
+ case "poisson" => if (isDefined(link)) Poisson($(link)) else
Poisson("log")
+ case "gamma" => if (isDefined(link)) Gamma($(link)) else
Gamma("inverse")
+ }
+
+ override protected def predict(features: Vector): Double = {
+ val eta = BLAS.dot(features, coefficients) + intercept
+ familyLink.fitted(eta)
+ }
+
+ @Since("2.0.0")
+ override def copy(extra: ParamMap): GeneralizedLinearRegressionModel = {
+ copyValues(new GeneralizedLinearRegressionModel(uid, coefficients,
intercept), extra)
+ .setParent(parent)
+ }
+}
+
+/**
+ * A description of the error distribution and link function to be used in
the model.
+ * @param link a link function instance
+ */
+private[ml] abstract class Family(val link: Link) extends Serializable {
+
+ /** Initialize the starting value for mu. */
+ def initialize(y: Double, weight: Double): Double
+
+ /** The variance of the endogenous variable's mean, given the value mu.
*/
+ def variance(mu: Double): Double
+
+ /** Weights for IRLS steps. */
+ def weights(mu: Double): Double = {
+ val x = clean(mu)
+ 1.0 / (math.pow(this.link.deriv(x), 2.0) * this.variance(x))
+ }
+
+ /** The adjusted response variable. */
+ def adjusted(y: Double, mu: Double, eta: Double): Double = {
+ val x = clean(mu)
+ eta + (y - x) * link.deriv(x)
+ }
+
+ /** Linear predictor based on given mu. */
+ def predict(mu: Double): Double = this.link.link(clean(mu))
+
+ /** Fitted value based on linear predictor eta. */
+ def fitted(eta: Double): Double = clean(this.link.unlink(eta))
+
+ /** Trim the fitted value so that it will be in valid range. */
+ def clean(mu: Double): Double = mu
+
+ val epsilon: Double = 1E-16
+}
+
+/**
+ * Gaussian exponential family distribution.
+ * The default link for the Gaussian family is the identity link.
+ * @param link a link function instance
+ */
+private[ml] class Gaussian(link: Link = Identity) extends Family(link) {
+
+ override def initialize(y: Double, weight: Double): Double = {
+ if (link == Log) {
+ require(y > 0.0, "The response variable of Gaussian family with Log
link " +
+ s"should be positive, but got $y")
+ }
+ y
+ }
+
+ def variance(mu: Double): Double = 1.0
+
+ override def clean(mu: Double): Double = {
--- End diff --
Here we constrict ```mu``` in valid range using a different method compared
with R. In R, if ```mu``` or ```eta``` is invalid, it will diminish
```coefficients``` until it makes ```validmu``` and ```validaeta``` passed. I
think is will make convergence slowness. I'm looking forward to hear others'
thought.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
